Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802.3EEESG] On the topic of tranistion time (was Re: [802.3EEESG] Comments on our work from Vern Paxson)



Wael,

I have sent mail to Michael and to Tony Jeffree, Chair of 802.1, 
informing them of our work and our desire for their participation in EEE 
as we could use their expertise.  Good idea to ask for a tutorial on AVB 
protocols.

Thanks,

Mike

Wael Diab wrote:
>
> This is a very interesting discussion, and an important one I think. 
> May I suggest that in July we send a request from our group to the AV 
> Bridging group in 802.1 to give us a tutorial presentation on both 
> their reservation and timing protocols as well as any constraints they 
> may have...at their earliest convenience ofcoarse.
>
> I think it may be helpful for those of us not intimately familiar with 
> the technology and for those of us that are familiar with it but may 
> have some additional questions.
>
> my 2c
> Wael
>  
> P.S. I have copied Michael who is Chairing the Audio/Video Bridging 
> Task Group
>
> --
> Wael William Diab
> Director, Technical Strategy, Broadcom
> Secretary, IEEE 802.3 CSMA/CD Working Group (Ethernet)
>
> -----Original Message-----
> From: Geoff Thompson [mailto:gthompso@NORTEL.COM]
> Sent: Monday, July 02, 2007 8:43 AM
> To: STDS-802-3-EEE@listserv.ieee.org
> Subject: Re: [802.3EEESG] On the topic of tranistion time (was Re: 
> [802.3EEESG] Comments on our work from Vern Paxson)
>
> Vern-
>
> At 11:51 PM 7/1/2007 , Vern Paxson wrote:
> > > It was PRECISELY my point that 802.1 AVB is not "designed for
> > > general Internet use".
> >
> >Then it certainly makes sense to avoid transition outages when it's 
> in use.
> >
> >Idon't know anything about 802.1 AVB.  At least naively, it seems there
> >will be signaling involved, such that switches can tell when it's being
> >used, and avoid transitions - is that right?
> >
> >                 Vern
>
> That was my thought. AVB enabled switches will be able to tell when it 
> is being used. I envisioned that the "stream coming" signal would be 
> more likely to come from the application than the switch though.
>
> There is a project that is part of AVB, P802.1Qat, Stream Reservation 
> Protocol to do this sort of thing for bandwidth reservation above the 
> PHY level..
> See: http://www.ieee802.org/1/pages/802.1at.html
>
> Geoff
>


-- 
Michael J. Bennett
LBLnet Services Group
Lawrence Berkeley Laboratory
Tel. 510.486.7913